TABLE Subcommand (PROXSCAL command)

The TABLE subcommand specifies the row identifier rowid and the column identifier columnid. Using TABLE, the proximities of separate sources are given in separate variables on the PROXSCAL variable list.

In the same manner, sources are identified by sourceid. In combination with rowid and columnid, the proximities are stacked in one single variable, containing the proximities of all sources, where sources are distinguished by the values of sourceid.

Using sourceid as the only variable on the TABLE subcommand indicates the use of stacked matrices, where individual stacked matrices are recognized by different values of sourceid.

  • Rowid, columnid, and sourceid should not be specified on the variable list.
  • When specifying both upper- and lower-triangular parts of the matrix, the SHAPE subcommand will determine the handling of the data.
  • If a cell’s value is specified multiple times, the final specification is used.
  • Rowid, columnid, and sourceid must appear in that order.
  • Omitting sourceid causes PROXSCAL to use the sources specified on the PROXSCAL variable list. Each variable is assumed to contain the proximities of one source.
  • Specifying multiple sources on the PROXSCAL variable list in conjunction with specifying rowid, columnid, and sourceid is not possible and causes PROXSCAL to ignore sourceid.

rowid. Row identifying variable. The values of this variable specify the row object of a proximity. The values must be integers between 1 and the number of objects, inclusive.

columnid. Column identifying variable. The values specify the column object of a proximity. The values must be integers between 1 and the number of objects, inclusive.

sourceid. Source identifying variable. The values specify the source number and must be integers between 1 and the number of sources, inclusive. The value labels of this variable are used to identify sources on other subcommands. These value labels must comply with variable name conventions. Omitting a value label causes PROXSCAL to use the default label SRC_n, where n is the number of the source.

Example 

DATA LIST 
  /r_id c_id men women. 


BEGIN DATA 
2 1 1.08 1.14 
3 1 0.68 1.12 
3 2 0.95 0.75 
4 1 0.96 0.32 
4 2 0.76 0.98 
4 3 0.47 0.69 
. . .... .... 
.. .. .... .... 
13 10 0.55 0.86 
13 11 0.61 0.97 
13 12 0.46 0.83 
END DATA. 


PROXSCAL men women
  /TABLE=r_id BY c_id
  /PLOT = INDIVIDUAL (women).
  • PROXSCAL reads two proximity matrices (men and women), where the row objects are specified by r_id and the column objects by c_id.
  • A chart of the individual space for women is plotted.

This is one way to proceed. Another way is to add the proximities of the additional source below the proximities of the first source and specify sourceid on the TABLE subcommand, containing values distinguishing the first and the additional source (see the next example).

Example 

DATA LIST 
  /r_id c_id s_id prox. 


BEGIN DATA 
2 1 1 1.08 
3 1 1 0.68 
3 2 1 0.95 
4 1 1 0.96 
4 2 1 0.76 
4 3 1 0.47
. . . .... 
.. .. . .... 
13 10 1 0.55 
13 11 1 0.61
13 12 1 0.46
2 1 2 1.14
3 1 2 1.12
3 2 2 0.75
4 1 2 0.32
4 2 2 0.98
4 3 2 0.69
. . . ....
.. .. . ....
13 10 2 0.86
13 11 2 0.97
13 12 2 0.83
END DATA. 


VALUE LABELS s_id 1 ‘men’ 2 ‘women’.


PROXSCAL prox
  /TABLE=r_id BY c_id BY s_id
  /PLOT = INDIVIDUAL (women).
  • PROXSCAL reads two proximity matrices. The row objects are identified by r_id and the column objects, by c_id. The proximity matrices are gathered in one variable, source01, where each source is distinguished by a value of the source identifying variable s_id.
  • A chart of the individual space for women is plotted.

Example

DATA LIST LIST
  /obj_1 obj_2 obj_3 obj_4 s_id


BEGIN DATA
0  0  0  0 1
1  0  0  0 1
2  3  0  0 1
4  5  6  0 1
7  0  0  0 2
0  0  0  0 2
8  9  0  0 2
12 11 12 0 2
END DATA.


VALUE LABELS s_id 1 ‘women’ 2 ‘men’.


PROXSCAL obj_1 obj_2 obj_3 obj_4
  /TABLE = s_id
  /MODEL = WEIGHTED
  /PLOT = INDIVIDUAL (women).
  • PROXSCAL reads two proximity matrices. The objects are given on the PROXSCAL variable list. Each source is distinguished by a value of the source identifying variable s_id, which is also used for labeling.
  • A chart of the individual space for women is plotted.